Thanks for sharing your critique! I enjoy a lively debate about evidence our interpretation of it. I didn’t write my post as a tear down of Stuart’s claims, but since you’re interest in that, let’s go through each of your claims.
Yes, Reges acknowledges sociocultural factors early in the article, and discusses his own efforts to change culture to make CS more welcoming. He then goes on to vaguely dismiss — ignoring all evidence — stereotype threat and imposter syndrome as critical factors in interest development. He continues, saying, “Certainly there are bad actors and companies where the culture is broken, but the vast majority of women work at companies that make significant efforts to provide a supportive work experience”, dismissing culture as anything more than a minor factor in all but the worst cultures. Is culture a factor or isn’t? I can’t tell. He seems to believe in some non-sociocultural, gendered notion of interest that is maybe biological or maybe something else. He doesn’t say, and he seems dismissive of culture (especially the culture of his own classrooms) as a factor, and so I infer his position as refuting culture as a significant factor.
Next, you suggest that the correlations explores in the Atlantic article are support for Reges’s claims, and that research on interest development are not. Here’s the critical idea you’re missing about both studies: the article in the Atlantic study is not about what causes interest and does not (and cannot with its data) control for the factors that cause interest. The article I cite, which is a synthesis of hundreds of studies on interest development, demonstrates that choices women (and men and everyone in between) make about what interests they pursue are far from determined by gender: they are the culmination of complex social processes, which are themselves gendered, leading to differences in which interests people of different genders develop. Culture causes gendered experiences, partially contributing to interest, which influence choice.
Take my mother for example, who I asked today, “why did you decide to become a teacher?” Her answer was that in the small farming community she grew up in, it was clear her socially acceptable options were wife, secretary, nurse, or teacher. She didn’t want to be a wife, she hated blood, and a teacher had told her she was terrible at typing. So she became a teacher. All this while black female mathematicians worked at NASA to get us to the moon, in another world she did not know. This “choice” was not a choice between all possible interests. It was a product of her culture, and her economic options. Stuart’s position appears to be that she just didn’t choose to be a mathematician because she’s a woman, and women have some innate predisposition to disliking math that scientists have not yet discovered.
You’ve framed my post as moralizing diversity and dismissing unpopular opinions. The reality is that Stuart, a highly experienced teacher with little expertise on diversity, and no expertise on human development or educational or social psychology, overlooked leading science on interest development that has clearly shown that people do not “choose” their professions: they are socialized into them through the complex interactions of their experiences, their communities, and their economic opportunities. That was sloppy.
I looked closely for compelling evidence in his article to falsify our leading theories about interest. Nothing he presented, however, was about interest or it’s causes, only by-products of it, such as choice of major. Until he presents data showing that interest is an innate construct, somehow tied to sex, gender, gender identity, or gender presentation, and not a socially constructed thing, there’s nothing really to unseat our most predictive explanatory theories. (Let’s not get into the discussion of how gender, is partly socially constructed too).
All that said, I’m a scientist. Show me some of us have a CS interest gene and I’ll be the first to discard our leading theories.